SDS 564: Clem Delangue on Hugging Face and Transformers

Initial Steps

Clem Delangue advises companies new to machine learning to start with pre-trained models available on the Hugging Face hub. He suggests beginning with simple features like classifying customer support emails or building autocomplete functions. This approach helps organizations build their 'machine learning muscle' progressively.

Start with something really simple, like you have customer support emails. How do you classify that? How do you extract information from that?

--- Clem Delangue

Delangue also highlights that the biggest roadblocks are often human-related, such as changing mindsets to accept the less deterministic nature of machine learning 1.

   

Ethics

Transparency and ethics are crucial in machine learning, according to Clem Delangue. He recommends using tools like model cards to clearly communicate what a model can and cannot do, both functionally and ethically. This transparency helps mitigate concerns about bias and misuse.

You can use tools like Gradeo, for example, which is an easy way to create demos for your machine learning models so that other people can try with their own examples and see if it works or if it doesn't work.

--- Clem Delangue

Delangue emphasizes the importance of using machine learning appropriately within workflows to ensure ethical use 2.